Objective 1: Create novel molecular interaction models of cell control networks relevant to cancer therapeutics using the molecular interaction map (MIM) graphical notation. The molecular interactions that govern how cells decide whether and when to divide have recently been elucidated in great detail. To grasp the complexities and functional possibilities of these networks, however, suitable diagrammatic representations are needed. The MIM notation with hierarchical map assembly (Kohn KW, et al., Cell Cycle 8: 2281-99, 2009) that we developed is arguably the best way to accomplish this goal in a manner that includes the essential details of the interactions. The task, however, must be divided into manageable sections that can then be integrated. We are developing a MIM of the role of beta-catenin in the regulation of cell proliferation, including the inputs from various receptors at the cell surface, the coordinate role of the actin cytoskeleton, the beta-catenin degradation machinery, the switching controls due to changes in beta-catenin phosphorylation states, and beta-catenin-dependent controls on gene expression in the nucleus. Information for the component interactions will come from current literature and from correspondence with authors. We aim to develop a comprehensive and comprehendible picture of this complex system. We recently extended this project based on clues derived from the Laboratorys analyses of NCI60 drug sensitivity data, which pointed to the determinative importance of the interactions of integrins and epidermal growth factors. This immediately led us to the network that controls integrin-based cell migration, cadherin-based cell-cell adhesions, and the epithelial-mesenchymal transition (EMT). Our previous mapping of the interactions of beta-catenin is an important component of this broadened scope. Recent work from many labs has substantially elucidated the molecular interactions controlling the EMT. Of particular interest with respect to the development of epithelial malignancies and potential for therapy are the recent advances in our understanding of the apoptotic event consequent to loss of cell-cell contact, a process known as anoikis. The spread of cancer cells requires these cells to have acquired resistance to the natural anoikis mechanism. New opportunities for therapy have been proposed based on overcoming this resistance. We have embarked on a project to assemble current data about the molecular networks that control this process and to map these networks using the MIM notation. At a minimum, the outcome will be an organized graphical portrayal of up-to-date information on the molecular networks controlling anoikis and the EMT. We also hope to propose, based on the comprehensive perspective provided by these MIMs, new insights into the function of the relevant control systems and to suggest new strategies for therapy. Objective 2: Answer the question: Can gene expression patterns of human tumor cells identify vulnerabilities that particular tumor cell types may have that would make them vulnerable to targeted therapy? We are analyzing the existing data on relative mRNA levels in 5000 genes from the NCI60 panel of human tumor cell lines. We apply clustering techniques to identify functional gene groupings that are co-expressed. We then assemble the available data on the roles these genes play in cell function and organize that information into a network model described by a MIM. This work is also linked to objective 1 (see above) on a gene expression/function cluster that appears to be focused on the EMT and related functions. Additionally, we are examining the gene expression patterns that are specific to melanomas. We have found that many of the expression patterns form subgroups that are identified with various aspects of melanosome development, as well as linkage to cells of the nervous system. We have embarked on a similar investigation of genes that are selectively expressed in colon cancer cell lines. Objective 3: Develop a drawing tool and plug-in for MIM, based on PathVisio. Currently, MIMs are drawn manually using standard graphics software (e.g., Canvas, Illustrator). This is laborious and these difficulties have impeded the widespread use of the MIM notation. Therefore we are developing a convenient drawing tool for MIM by modifying Pathvisio, an existing software. The MIM-modified Pathvisio will allow the creation and editing of MIM diagrams. Users will be able to create MIM elements in a way that is similar to other graphical editing software. Via interaction with the MIM API (see below), Pathvisio-MIM will guide users to use symbols consistent with the MIM grammar outlined in the MIM ontology (see below). The software will reduce the time required to learn the MIM notation by providing graphical elements specific to MIM, monitoring the validity of user constructs, and providing functionality that is typical for diagramming editors. The MIM graphics editor will maintain connections between diagram elements as these elements are moved. Implementation of Pathvisio-MIM will initially be based on the MIM notation definitions described in Kohn KW, et al. (Mol. Biol. Cell 17: 1-13, 2006). Future versions will add features as required. Objective 4: Develop formal specifications and exchange format for MIM. To provide a standard exchange format for MIM, we are developing a MIM ontology that will describe a data structure for the information contained in MIM diagrams (e.g., the relationships between the various graphical elements). This ontology will describe the objects and interactions in a textual format (graphics visualization will be added later using a separate software). The MIM ontology will share a format and some conventions that are currently used by BioPAX. The ontology will provide a grammar for storing information about MIM interactions based on widely used computer standards. The result will be a well-defined and complete description of the elements for implementing the MIM notation. A written specification will complement the ontology for some needed rules that are outside the scope of the ontology format itself. The ontology and its specification will provide a textual basis describing the MIM notation for MIM software development. The ontology and specification will be implemented as an application programming interface (API), providing a foundation for developing MIM-related software. By itself, the MIM API will determine how MIM-related software can process the information content represented by MIM diagrams, independent of its diagrammatic visualization. Using the MIM API, software can then be developed to export and/or import MIM files to/from other textual formats and analysis tools. Later, the MIM API will be coupled to software components that are related to the visual aspects of MIM diagrams. Objective 5: Test hypotheses about the role of MdmX in the regulation of p53, based on our previous simulation study. Our computer simulation studies (Kim S, et al., PloS Comput. Biol. 6: e100062, 2010) suggested previously unknown roles for MdmX in the system that controls the function of tumor suppressor p53. We are now testing these predictions by quantitative kinetic studies in cell culture, asking whether MdmX amplifies, stabilizes, and/or confers the switch-like behavior suggested by our simulation results. We will then use the quantitative data from these experiments to refine our computer models.